Like any new technology, the winners and losers are yet to be determined, however leaders are emerging.
Voco has been working with three clients lately where the evaluation of SD-WAN options has come up. In each case they had a large, international MPLS network, some aging WAN optimisation and numerous legacy business applications delivered from a few traditional data centres. All three clients wanted better price/performance and wanted to make cost savings.
In referring to SD-WAN I am talking about the core technology set that enables a cost and operationally-efficient hybrid WAN through a central orchestration platform that controls multiple endpoint devices. While hybrid WANs have been around for some time, the addition of fully centralised deployment and management capability promises cost benefits by simplifying the task of mixing traditional WAN and local internet access services into a complete hybrid WAN outcome.
Like any new technology, the winners and losers are yet to be determined, however leaders are emerging. In the case of SD-WAN the proprietary nature of the integration between the central orchestration engine and the endpoint devices of each vendor’s solution means swapping out a vendor is an ‘all or nothing’ exercise, i.e. you can’t mix and match endpoint and orchestration providers.
That technology risk could be mitigated through consuming SD-WAN as a service. The wider risks to realising an overall cost saving, however, are not so easily outsourced. What would prevent SD-WAN simply being another cost on top of the current landscape? Before we get to that, let’s summarise the key benefits of SD-WAN:
In a nutshell, this solution combines instant hybrid WAN for flexibility and resilience with the cloud-like gains of an evergreen service and dramatically simplified deployment/onboarding. Sounds awesome? It is. But back to the cost. SD-WAN will attract extra charges on top of the underlying connectivity.
Of course, multiple diverse broadband links into your sites can be bought ‘as cheap as chips’. But that’s still a new cost to add into the mix, so where are the savings going to come from? The old MPLS network – those monthly charges – and potentially by avoiding some life cycle refresh that you have been deferring?
This is where we hit the key dependency issue every time – the legacy business applications. How will they perform over SD-WAN? And what is the process to migrate some to the cloud or replace them with SaaS based alternatives? Will this allow us to shrink the MPLS connections and/or coverage, i.e. can you drop any connections or are you stuck with the minimum size to all sites?
Essentially there are three key elements to assess when building a case for SD-WAN. They are represented graphically below. Your readiness for SD-WAN adoption and ability to construct a positive financial business case will be largely determined by your alignment across them.
Our three clients all decided they wanted some SD-WAN. In two cases, however, they want it in the future, so it currently sits on their technology roadmap until their application refresh plans catch up. In the third case they are moving ahead with a set of pilot sites to see how the technology performs with their remaining legacy applications. That will in turn confirm, or kill, their business case.
Are you facing user experience and/or cost challenges caused by your networking layer? Give us a call to discuss how this can be attacked from all fronts – application refresh/rationalisation, cloud/SaaS migrations and networking technologies like SD-WAN.
To discuss the topic further, please get in touch with the author, Selwyn Rimmer, on 027 224 4649.